Opposes: this is a consumer-protection probe, not an AI safety one. A senior FTC official says it turns on deceptive or unfair practices under the FTC Act, and Chair Ferguson says audit trails showed the AI followed instructions and backs no new rules.

USA Today's senior FTC source says the probe began before the rogue-agent incidents and will turn on whether the companies engaged in deceptive or unfair practices. Chair Andrew Ferguson, as reported by Reuters via Ars Technica, says reviews showed the systems were carrying out instructions, and he agrees with Trump that no new regulation is needed. That makes it a case about what the companies told consumers.
Benjamin Toby
Benjamin Toby·1mo

The FTC's own framing is narrower than "AI safety harms." A senior FTC official told USA Today on Sept 30 that the probe was launched before the recent rogue-agent incidents and "will turn on whether the companies engaged in deceptive or unfair business practices that violate the FTC Act." That is the agency's standard consumer-protection test, the same one it applies to any product.

The chair is sceptical of the safety narrative. According to Reuters, as summarised by Ars Technica, Andrew Ferguson said AI companies had sometimes described systems as acting beyond human control, but "subsequent reviews of audit trails have shown that the systems were carrying out instructions they had been given." He said: "if someone tells a tool to do something, and the tool does it, I don't think we would say, 'Oh, what do we do about the tool?'" Ars reports he aligns with Trump that no new regulations are needed.

On this reading, the question is whether marketing and safety claims misled the public, for example about how autonomous or secure the agents are. USA Today reports Ferguson accusing AI firms of whipping up panic to win rules that "build a moat around their businesses."

Limits. The demands do ask about security incidents and safeguards, and consumer harm from unsafe products is a safety harm. The FTC has not published the demands, so their full scope is known only through sources.

Exhibits

Evidence cited

USA Today, Sept 30, 2026: Is AI dangerous? FTC escalates major probe into OpenAI and Anthropicusatoday.com
Under review
Ars Technica: Report: FTC probes Anthropic, OpenAI after Trump dismissed AI fears as 'hoax'arstechnica.com
Under review

Discussion

Log in to join the discussion.
No comments yet. Start the discussion.